Software increases computational performance

Edited by the Engineeringtalk editorial team Jun 7, 2004

The latest release of Matlab enhances programmer productivity with new tools that enable rapid, iterative program creation.

The latest release of Matlab enhances programmer productivity with new tools that enable rapid, iterative program creation.

In addition, Matlab 7 offers built-in support for integer and single-precision floating-point maths, as well as language features for managing and analysing larger data sets.

A large number of optimisations across data types, operations, functions, and hardware result in improved computational performance of end user applications.

Also new in the Matlab family of products is an enhanced Matlab Compiler that now supports the full Matlab language, enabling developers to deploy many more Matlab applications for use outside of Matlab.

"Matlab 7 builds on the solid base of technical computing capabilities that The MathWorks has established with its product suite built around Matlab and Simulink products",said Paul Zorfass, Senior Analyst with IDC/FTI.

"The software's extended capabilities make it practical to address more applications such as fixed-point, geospatial and RF analysis".

"With an enhanced compiler in the Matlab family, this new version also meets the technical computing needs of individuals whose tools and data can be shared across workgroups and organisations".

Among its new programmer productivity capabilities, Matlab 7 features a redesigned desktop that includes new programming and debugging tools, automated program analysis for code quality, and publishing of code and output directly to HTML and Word.

New interactive point-and-click plotting tools enable faster and easier creation of Matlab graphics, including the ability to generate code that recreates plotting sessions for later use or redistribution.

Adding to its robust support for double-precision mathematics, Matlab now supports integer and single-precision data types and arithmetic.

Operating on data as integers or single-precision values instead of double-precision values offers significant savings in performance and storage, thereby enabling users to work with larger data sets.

This capability combined with the incorporation of new FFT algorithms enables Matlab users to perform a single-precision FFT 20% faster on average.

Also, Matlab now takes advantage of the Intel MMX library, resulting in up to eight times speed up for computations on integer data.

Available in the Matlab 7 family of products, Matlab Compiler now supports the full Matlab language, including most Matlab toolboxes.

The enhanced Matlab Compiler enables engineers to deploy their Matlab work to others as self-contained applications, or incorporate their algorithms in environments such as Excel, C, C++ and COM.

As a result, engineers and scientists can leverage the Matlab Compiler to deploy significantly more Matlab applications and enable companies to take advantage of Matlab technical capabilities throughout their organisations.

"This latest release of Matlab and the Matlab Compiler delivers the power, speed, and versatility engineers demand for their analysis, development, and deployment", said Jim Tung, MathWorks Fellow, The MathWorks.

"Matlab is the standard technical computing platform across a broad range of industries from aerospace and automotive to biotech and communications".

"With the significant features and capabilities delivered in Matlab 7, we're providing our customers with the tools to tackle their increasingly large and broad technical challenges now and into the future".

Matlab 7 is available immediately for Windows, Unix/Linux, and Macintosh systems.

Pricing starts at GBP 1650.